The Ontogeny of Information Susan Oyama ScienceandCulturalTheory ASeriesEditedbyBarbaraHerrnsteinSmith andRoyE.Weintraub The Ontogeny of Information Developmental Systems and Evolution SecondEdition,RevisedandExpanded Susan Oyama ForewordbyRichardC.Lewontin Secondedition,revisedandexpanded ©byDukeUniversityPress OriginallypublishedinbyCambridgeUniversityPress Allrightsreserved PrintedintheUnitedStatesofAmericaonacid-freepaper(cid:7) DesignedbyC.H.Westmoreland TypesetinTimesRomanwithDinNeuzeitdisplay byTsengInformationSystems,Inc. LibraryofCongressCataloging-in-PublicationData appearonthelastprintedpageofthisbook. Contents ForewordbyRichardLewontin vii PrefacetoSecondEdition xvii Preface xix 1 Introduction 2 TheOriginandTransmissionofForm: TheGeneastheVehicleofConstancy 3 TheProblemofChange 4 VariabilityandOntogeneticDifferentiation 5 VariationsonaTheme: CognitiveMetaphorsandtheHomunculoidGene 6 TheGhostsintheGhost-in-the-MachineMachine 7 TheOntogenyofInformation 8 Reprise 9 Prospects AfterwordtoSecondEdition Notes References IndexofNames IndexofSubjects This page intentionally left blank Foreword The first edition of The Ontogeny of Information () was instigated byagrowingimpoverishmentofexplanationandunderstandingofhow organismscometobe.Asaconsequenceofalonghistoryofdiscovery in genetics and especially in molecular genetics, together with a virtu- allycompletestagnationoftheclassicalscienceofexperimentalembry- ology,explanationsofhoworganismsgrowanddifferentiatetoproduce theircharacteristicformsandpropertieshadbecomeframedentirelyin terms of the ‘‘turning on’’ and ‘‘turning off’’ of genes. Already in , at the Cambridge University symposium commemorating the one hun- dredthanniversaryofDarwin’sdeath,SydneyBrennerinhiskeynotead- dresshaddeclaredthatgiventhecompletesequenceofanorganism and a large enough computer it would soon be possible to compute the organism.By,thegeneticboaconstrictorhadtotallyenfoldedem- bryologyinitshelicalcoils.Now,nearlyfifteenyearslater,thehelpless victimhasdisappeareddownthemolecularmawandisslowlybeingdi- gested.Intheend,ofcourse,theresultwillbewhatistobeexpectedfrom suchaprocess:aslightenlargementofthebodyoftheconsumerandthe productionofalargeamountoffecalmatter. AsSusanOyamaexplainsinTheOntogenyofInformation,therearetwo difficultieswiththecurrenttrendofexplanationofdevelopment.Thefirst isthatthewrongquestionisbeingasked;thesecondisthat,eventothe extentthatweareconcernedwiththeanswertothatquestion,thewrong answerisbeinggiven.Whenthewrongquestionisbeingasked,itusually turns out to be because the right question is too difficult. Scientists ask questionstheycananswer.Thatis,itisoftenthecasethattheoperations ofasciencearenotaconsequenceoftheproblematicofthatscience,but thattheproblematicisinducedbytheavailablemeans.ThegeneticistAl Hersheyusedtosaythatheavenisfindinganexperimentthatworks,and thendoingitoverandoverandover.Sequencingiseasy;anydamn viii Foreword foolwithamachinecandoit.Itisrathermoredifficulttomakegenetic constructsofjusttherightsorttoalterorknockoutthereadingofparticu- lar genes, but a competent Drosophila geneticist can usually manage it. Ontheotherhand,nooneseemstoknowevenhowtoaskanexperimen- talquestionabouttheactualcomingtogetherofmoleculestomakecell organelles,orwhymyrearappendageslooklikelegsandfeetwhilemy frontappendagesarearmsandhands,orwhyaparticularmicroanatomy ofthecentralnervoussystemallowsthereadertounderstandthispage. Thereasonthatthewronganswerisbeinggiven,eventothequestions weknowhowtoask,isthatsomeformsofthinkingwedependonlead usastray.First,wedonotknowhowtodispensewithmetaphoricallan- guageandareoftenvictimizedbypropertiesofthemetaphor.Physicists canspeakofthe‘‘billiardball’’modelofcollidingmoleculeswithoutsup- posingthatmoleculesaremadeofplastic,orthattheyareredandwhite, or that they make a clicking sound when they hit each other. But when biologistsspeakofgenesas‘‘computerprograms,’’theyerroneouslysup- pose that all the organism’s attributes are prefigured in its genes and all thatisrequiredforafixedoutputisfortheenterkeytobepressed. Second,unlikephysicists,whohavelearnedtoberathermoresophis- ticated, biologists, who imagine that theyare following the example of physics, have a simplistic notion of causes and effects. If A and B are causallyconnected,theythink,theneitherAisthecauseandBtheeffect or viceversa. But no physicist would try to identify the separate cause andeffectinthecoupledelectromagneticequations.Themagneticfield does not cause the electric field any more than the electric field causes themagnetic.Thereissimplyapairofcoupleddifferentialequationsde- scribingtheelectromagneticphenomenathataretwoaspectsofthesame process.Norisitthecase,asbiologistsusuallybelieve,thatequalcausal forces have equal effects. In a dynamical process in time it is also nec- essary to take account of initial conditions. If I drive eighty miles west andtwenty-fivemilesnorth,itisimpossibleforyoutoknowthatIwill arriveinBrattleboro,Vermont,unlessyouknowthatIstartedinBoston. Norcanphysicalprocessesbecountedontobehaveinadeterminateway whentheyarebasedonaverysmallnumberofreactivemoleculesirregu- larlydistributedinspace,asisthecaseformostoftheimportantreactants withincells. Throughoutthehistoryofmodernbiologytherehasbeenaconfusion Foreword ix betweentwobasicquestionsaboutorganisms:theproblemoftheoriginof differencesandtheproblemoftheoriginofstate.Atfirstsighttheseseem tobethesamequestion,andtakenintherightdirection,theyare.Afterall, ifwecouldexplainwhyeachparticularorganismhasitsparticularform, thenwewouldhaveexplained,paripassu,thedifferencesbetweenthem. But the reverse is not true. A sufficient explanation of why two things are different may leave out virtuallyeverything needed to explain their nature.SupposethatontherailwaylinebetweenBostonandNewYork there is a switch that, when set to the right, sends the train on a direct southwesterlyroutebetweenthecities,but,whensettotheleft,shuntsthe trainonacircuitouspathwestwardandthensouthwardalongtheHudson River.Inresponsetothequestion,‘‘Whydidittakeyoueighthourstoget toNewYorkfromBostoninsteadoftheusualfive?’’Imightquiteprop- erlyanswerthattheswitchwassettotheleft,butthattellsnothingabout theimmensecomplexityofmechanicalandsocialforcesinvolvedinthe phenomenon of train travel. Indeed, the scheme of explanation of form offeredbymoderndevelopmentalgeneticsisthatofasignalingnetwork amonggenesthatresemblesnothingsomuchastheilluminatedmapon thewallofarailroadcontrolroomwithredandgreenlightstwinklingon andoffastheswitchesarereset.Thetemptationtosubstitutetheexpla- nationofdifferencesforanexplanationofoutcomes,substitutingamap of the switches foran articulated physical and social explanation of the movementsoftrains,hasplaguedbiologyatleastsincethemiddleofthe nineteenthcentury. The experiments of Mendel on variation in peas were not motivated simplybyidleintellectualcuriosity;theywerepartofaresearchprogram motivatedbyfruitbreedinginMoravia,aprogramtodiscoverthequan- titativelawsofvariationinplants.Mendel’spaperwasentitled‘‘Experi- mentsonPlantHybrids,’’andhetellsusinthe‘‘IntroductoryRemarks’’ thatamongpriorinvestigationsbyothers,‘‘notonehasbeencarriedout tosuchanextentandinsuchawayastomakeitpossibletodeterminethe numberofdifferentformsunderwhichtheoffspringofhybridsappear, ortoarrangetheseformswithcertaintyaccordingtotheirseparategen- erations,ordefinitelytoascertaintheirstatisticalrelations.’’Notaword here about the origin or inheritance of form itself. These were not ex- perimentsmeanttoexplainwhysomeflowersareredandotherswhite, whysomeplantsaretallandothersshort.Hewasapparentlyconcerned x Foreword onlywiththe‘‘statisticalrelations’’ofdifferences.ItwasMendel’ssingle- minded concentration on variation, on the statistical regularities of dif- ferencesarisingintheoffspringofhybrids,thatwashisepistemological breakfromthepastwhichmadepossiblethescienceofgenetics.Yet,at theverymomentofhisradicalmovefromstudyingsimilaritiestostudy- ing differences, Mendel himself slipped over into an explanation of the originofformitselfbypostulatingthatthered-andwhite-floweredplants hadmysterious‘‘factors’’forrednessandwhiteness.Moreover,thefactor forrednesshadapowerthatenabledittodominatethefactorforwhite- ness when the twowere present in the same plant. The language of the originalproblematicandofthedescriptionoftheresults,alanguageen- tirelyofdifferences,becomes,intheexplanatoryhypothesis,alanguage oftheoriginofform.Soatitsverybirth,thestudyofthelawsofvaria- tioninhereditypreparedtobethestudyoftheheredityoftype.Genetics preparedtoswallowanddigesttheproblematicofembryology.Thesub- sequenthistoryofgeneticsisinlargepartthestoryofthisingestion. Duringthefirsthalfofthepresentcenturythemajorgoalofgenetics, inadditiontotheelaborationoftherulesofassortmentofgeneticdiffer- ences in heredity, was to uncover the material basis of which the rules arethequantitativemanifestation.Butthissearchforamaterialbasishas beenakindofreinforcingpracticeforthetheoreticalviewthattheexpla- nationofdifferencesisthesameastheexplanationofoutcomes.Thisis seeneveninthenamingofgenes.Geneticistsspeakcasuallyofthe‘‘gene forwhiteeyes,’’but,ofcourse,thereisnosuchgene.Thereisavariety ofgeneswhosereadingbythecellisproximallyinvolvedintheproduc- tionofeyepigmentanditsdepositionintheeyecells.Oneofthese,the so-calledwhitelocus,canbesoalteredfromitsusualformbymutations thatnoeyepigmentcanbedeposited.Thisgeneisneitherthegene‘‘for’’ white eyes nor the gene ‘‘for’’ red eyes, and geneticists know that very well,yettheconstantrepetitionofthelanguageofgenes‘‘for’’whiteeyes or‘‘for’’vestigialwingshasapowerfuleffectontheconceptualizationof thedevelopmentalprocess.Hobbes’swittyremarkthat‘‘wordsarewise men’scounters,theydobutreckonwiththem,buttheyarethemoneyof fools’’failstotakeaccountoftheinevitablefetishismthatarisesfromthe constantuseofcounters. Geneticists’concentrationontheoriginofdifferences,andtheircon- fusionofthisquestionwiththeprocessesleadingtoanorganism’sstate,